backend-dev-guidelines
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFE
Full Analysis
- [SAFE] (SAFE): The skill consists of documentation and code templates meant to guide developers. It does not contain any executable scripts, tool definitions, or automated actions that would pose a risk to the agent's host environment.- [CREDENTIALS_UNSAFE] (SAFE): The configuration guidelines explicitly advise against hardcoding secrets and provide clear patterns for using environment variables and
.gitignoreto protect sensitive information. Placeholder values (e.g., 'your-secret-here') are used appropriately in examples.- [DATA_EXFILTRATION] (SAFE): The skill includes security-conscious code patterns for error monitoring, specifically instructing developers to scrub sensitive headers like 'authorization' and 'cookie' and to mask email addresses for PII protection before data is sent to external services like Sentry.- [PROMPT_INJECTION] (SAFE): No instructions were found that attempt to override the AI agent's system prompt, bypass safety filters, or disclose internal instructions.
Audit Metadata